Research outline

Seismic Motion Measured in Mountain Tunnel

Photo 1 Mountain tunnel damaged by
the Niigata Prefecture Chuetsu
Earthquake (photo by the Road
Department, Hokuriku Regional
Development Bureau, MLIT)

Fig. 1 Location of the tunnel in
which measurement equipment was
installed, and the epicenter
of the measured earthquake

Fig. 2 Photo showing installation
positions of measurement devices

Fig. 3 Ground deformation and tunnel
damage modes (click to enlarge)

  Mountain tunnels have generally been constructed of earthquake-resistant structures, except for those constructed at extremely unfavorable locations. This conventional approach, however, was challenged by the Niigata Prefecture Chuetsu Earthquake in 2004. Some tunnels were so damaged that concrete chunks dropped from the tunnel lining, which is generally installed in Japan as an inmost concrete wall about 30 to 60 cm in thickness (Photo 1). To minimize such damages with carefully reasoned seismic measures, the mechanisms involved in earthquake damage to mountain tunnels must be elucidated.

  There are, however, only a small number of cases of mountain tunnels being damaged by earthquakes. In addition, there are an extremely small number of cases in which seismic motion was measured in an actual mountain tunnel. Therefore, we have almost no knowledge of a tunnel's behavior when undergoing major seismic motion. In light of this, we installed equipment in an existing tunnel to measure motion caused by earthquakes, and successfully obtained measurement data after a major earthquake actually occurred.

  Measurement equipment was installed in an ordinary two-lane road tunnel, the San Juan Tunnel (Fig. 1), constructed in 1996 in Ishinomaki City, Miyagi prefecture. The equipment, installed in the tunnel lining, included accelerators and strain gauges (Fig. 2). After installation, on Apr. 2011, a major earthquake occurred off Miyagi prefecture, with magnitude 7.1 (one of the aftershocks of the Great East Japan Earthquake), which shook Ishinomaki City with a seismic intensity of Level 6 according to the observation data.

  The tunnel acceleration caused by this earthquake was roughly 200 gal (roughly 0.2 times gravitational acceleration), somewhat less than the 300 gal measured at ground-surface measuring points elsewhere in the city. This may be evidence that less motion is felt beneath the ground than at the surface.

  The maximum level of strain occurring in the tunnel lining was about 20μ (deformation of a 10m long structure having shrunk by 0.2mm), far less than the 2,000μ lower limit of compressive strain sufficient to destroy concrete. In addition, the mesurement results suggest that the pattern of actual deformation was more similar to compressive deformation from the vertical or lateral direction (Type-II or III in Fig. 3); though the most prevalent form of deformation in the ground caused by an earthquake is typically shear deformation (Type-I of Fig. 3), and this is considered the most basic type of deformation when earthquake-resistant measures for a tunnel are reviewed. These results indicate the necessity to consider types of deformation other than shear deformation, when studying seismic measures for mountain tunnels in the future.

  Moreover, the measurement results may corroborate the conventional notion that tunnels are structurally resistant to earthquakes, and suggest new findings on tunnel deformation in the event of an earthquake. The measurements in question, however, were made for only a single section of a single tunnel. It is important to gather further data and reviews in the future, in order to confidently establish measures for seismic reinforcement of tunnels.

(Contact: Tunnel Research Team)

Danger Warning with Color and Light

 Morpho butterfly

 Metal strain detected by color change
(red to green).
Deformation in the green surface is shown.
(click to enlarge)

 Concrete cracking detected by
brightness change (brighter to darker).
(a) Cracks at the center are not
easily recognized under natural light.
(b) Under black light,
the same cracks are more visible.
(click to enlarge)

  If a structure could itself warn us that it is about to collapse, we would need far less inspection.

  Road structures such as bridges and tunnels, and river structures such as levees and water gates, have undergone massive destruction due to earthquake-induced damage or aging. However, we may be unable to perceive the danger of destruction (or deterioration) of these structures when we merely examine them from a distance. In order to make detailed observations we must establish a temporary working floor. In such detailed observation, we touch the structure or listen to the sound of tapping, in order to determine whether the structure is vulnerable to damage; however, some structures can stretch for dozens of meters, or even kilometers. Since such wide areas must be manually inspected, a great deal of time and manpower is required, and we can therefore miss signs of deterioration or danger.

  We are working on materials (functional materials) capable of informing us, through color or light, that the structure is under strain or has cracks because of some external force being applied. For example, a film called “opal film” changes color when deformed, thereby indicating the presence of strain. Opal film imitates the wings of a Morpho butterfly, which have brilliant colors and a metallic luster. The film exploits the fact that the coloring mechanism based on light interference caused by the microscopic structure of scales is affected by deformation. In addition, we have a functional material called sensor paint, which, when applied to concrete or steel, warns us through light of danger the instant cracking occurs. The luminescent feature is effective for inspecting dark and poorly accessible parts of structures, such as the inside of a tunnel or the underside of a bridge.

  If these functional materials are pasted or painted on a road or bridge, they can instantly tell us when the structure receives abnormal force or has minute cracks. Moreover, as structures to which such functional material is applied can be observed from a distance with a digital camera, inspectors can examine such structures without need of erecting a scaffold in a dangerous location. We intend to work toward practical use of these materials for the contribution to society. The research results reported here are the outcome of joint research by PWRI, the National Institute for Materials Science, Hiroshima University, and the Tokyo Institute of Technology.

(Contact: Advanced Materials Research Team)

Development of a Turbid Water Treatment System with Less Environmental Impact

Photo 1 Dam reservoir with turbid
water after heavy rainfall

Photo 2 Laboratory test for
coagulation and precipitation
(Top to bottom: after resting
for 1 min., 15 min., and 60 min)
(click to enlarge)

Fig. Change in turbidity
level for the two tanks
(click to enlarge)


  A dam stores water in its reservoir when heavy rain falls, then gradually releases the stored water downstream when the flood danger has disappeared, and we use such water for drinking and agriculture. Water flowing into a dam reservoir as a result of heavy rain is more turbid than ordinary river water (Photo 1). The turbidity of water released from such a reservoir can affect the downstream landscape or the habitat environment of aquatic life, and this may cause serious problems. PWRI is working on development of a technology for water turbidity improvement, as a solution for turbid water in dam reservoirs.

Treatment System Using Flocculants

  One of the methods to lessen water turbidity is coagulation and precipitation. When flocculants are mixed in turbid water and the water is agitated, the flocculants adsorb the soil particles that are the source of turbidity, and the soil particles form clusters, eventually promoting precipitation. As a result, the supernatant becomes transparent and clean over time (Photo 2). Chemical flocculants are generally used in this process. But if such chemicals are spread over a dam reservoir, recovery of the precipitates becomes a problem. As a solution to this problem, we designed a system that uses naturally derived flocculants to remove turbidity, a technique that causes no major environmental problems even though the flocculants are directly scattered over the dam reservoir.

  The flocculant used in our technique is an allophane, composed of clayey colloidal particles and present in great quantity in volcanic ash deposit. When an allophane is dissolved in water, and acid or alkali is added, it changes its condition, becoming either dispersed or coagulated, depending on the pH level. Exploiting this characteristic, we developed a turbidity-improving system involving the adsorption of soil particles, the source of water turbidity, and tested it under field conditions on an experimental basis.

Field Test

  A field test of turbidity improvement was conducted on the shore of the reservoir of the Yamasubaru Dam managed by Kyushu Electric Power Company. Removal of sediments deposited on the bottom of the reservoir is ongoing at the Yamasubaru Dam. Turbid water generated by this sediment removal work was pumped up into water tanks, where the coagulation test was conducted. Two water tanks were set up, one in which the turbidity treatment was conducted (Tank 1), and another in which no treatment was performed (Tank 2).

  The turbidity change in the water of the two tanks is shown in this figure. In this test, the start and end times of Tank 1 agitation were set at -180min. and 0min., respectively, and the change in turbidity over time, from -180 to 1,440min., was recorded. In Tank 1, ultrasonic vibration was administered to 6㎥ of turbid water for 3 hours, during which the water was continuously agitated. As a result, the turbidity level was remarkably reduced in comparison with Tank 2, which underwent no treatment. The turbidity level at the start of agitation was roughly 290NTU (the unit of turbidity) on average, in the upper and lower layers, for both Tanks 1 and 2; and the turbidity level decreased over time for both tanks. After the lapse of 1,440min., however, the average turbidity level in the upper and lower layers was 13NTU for Tank 1 and 186NTU for Tank 2, confirming that allophanes were effective in promoting coagulation.

Toward Practical Use

  Although the test water was almost entirely cleansed of turbidity, there are still numerous problems to solve before the system can be put to practical use. We intend to continue research toward practical use of the system by, for example, improving coagulation treatment efficiency.

(Contact: River and Dam Hydraulic Engineering Research Team)

How are snowfall and snow depth changing?

 Fig. 1 Trends of change in seasonal
maximum snow depth
(FY1983 — FY2008)
(click to enlarge)

 Fig. 2 Difference in frequency of heavy
snowfall events
(≧40cm for 24-h snowfall)
: avg. of FY2001 to FY 2010 vs
avg. of FY1981 to FY2000
(click to enlarge)

  Whereas some recent years have had little snowfall due to warm winters, other recent years had much snowfall, as seen in the heavy snowfall event of 2006 and the heavy winter snowfalls of 2010 and 2011. Furthermore, localized heavy snowfalls or snowstorms driven by rapidly developing low-pressure systems have caused vehicle-stranding disasters in various locations. One notable case was the snowstorm in Hokkaido in March 2013 that left nine people dead. Such unprecedented severe weather conditions have occasionally been observed.

  A survey focusing on Niigata Prefecture and northward regarding recent trends in maximum depth of snow cover and events of short-term heavy snowfall was carried out. The data used for this study were hourly snow depth figures at 141 points observed by the Japan Meteorological Agency's Automated Meteorological Data Acquisition System (AMeDAS) from 1981 (start of observation) to 2010. The winter period was defined as November 1 to April 30.

  Figure 1 shows the trends in maximum depth of snow cover per winter for the 26 winter seasons from 1983 to 2008. Seasonal maximum snow depth shows an increasing trend in some areas of Hokkaido, such as the Sea of Japan coastal regions north of Otaru, inland regions and the Sea-of-Okhotsk coastal region, and in some areas in Honshu such as the northern Pacific region.

  Next, trends of the occurrence frequency of heavy snowfall events (≧40cm for 24-h snowfall) are examined. In this examination, the hourly increase of snow depth was regarded as the hourly snowfall.

  Figure 2 shows how the occurrence frequency of heavy snowfall events in the last decade from 2001 to 2010 changed from the previous two decades (1981 to 2000). A tendency for the occurrence frequency of such events to increase can be observed mainly in Eastern Hokkaido and the mountainous areas of the Tohoku Region. The fact that the heavy snowfall events in Eastern Hokkaido were mainly driven by well-developed low-pressure systems implies that weather patterns which cause heavy snowfall from the development of a winter low-pressure system have increased. The Sea of Japan side of Honshu shows a tendency for the occurrence frequency of heavy snowfall events to decrease. The causes of this declining trend are presumed to include increases in winter temperatures that cause precipitation to fall as rain rather than as snow.

  Maximum depth of snow cover and maximum daily snowfall are used in snow control planning and in the design of snow protection facilities for roads. Analytical results on the trends of such data on snowfall and snow cover can be basic material for deliberating on snow control measures.

(Contact: Snow and Ice Research Team, CERI)